Prove2Me
Navigate
DiscoverCollectionsFormalpediaGalaxyBlogsUsersMomentumMy Missions+
Prove2Me
Log in
← Formalpedia

Theorem 4.1 — for disjoint constrained budgeted sets, z_Aff(U) = O(log² n / log log n)·z_AR(U)

Proved
BudgetedAffine.Disjoint.theorem_4_1

by mikedeng1 · Oct 13, 2026 · Mathlib 0df444a (Lean v4.33.1)

affine-policiesapproximationp2o-batch-pfp2bp2o-gran-per-chapterp2o-plan-paperp2o-v1robust-optimization

Consider the two-stage adjustable robust problem (1.1) with A∈Rm×nA\in\mathbb R^{m\times n}A∈Rm×n, B∈R+m×nB\in\mathbb R^{m\times n}_+B∈R+m×n​, c,d∈R+nc,d\in\mathbb R^n_+c,d∈R+n​, X\mathcal XX polyhedral with cTx≥0c^Tx\ge0cTx≥0 for all x∈Xx\in\mathcal Xx∈X, and U\mathcal UU the disjoint constrained budgeted set (4.1) for a partition S1,…,SLS_1,\dots,S_LS1​,…,SL​ of [m][m][m] and weights wℓi∈[0,1]w_{\ell i}\in[0,1]wℓi​∈[0,1]. Then

zAff(U)=O(log⁡2nlog⁡log⁡n)⋅zAR(U),z_{Aff}(\mathcal U)=O\Big(\frac{\log^2n}{\log\log n}\Big)\cdot z_{AR}(\mathcal U),zAff​(U)=O(loglognlog2n​)⋅zAR​(U),

i.e. there are a constant C>0C>0C>0 and a threshold NNN such that for every n≥Nn\ge Nn≥N, every mmm, every number of blocks LLL and every such instance,

zAff(U)≤C⋅log⁡2nlog⁡log⁡n⋅zAR(U).z_{Aff}(\mathcal U)\le C\cdot\frac{\log^2n}{\log\log n}\cdot z_{AR}(\mathcal U).zAff​(U)≤C⋅loglognlog2n​⋅zAR​(U).

Here nnn is the number of decision variables. The bound does not depend on mmm or on the number LLL of budget constraints.

Formalization Note CCC and NNN are chosen before mmm, LLL and the instance. Values are extended reals, so an infeasible problem gives zAR=+∞z_{AR}=+\inftyzAR​=+∞ and the bound holds trivially. log⁡\loglog is the natural logarithm and log⁡2n=(log⁡n)2\log^2n=(\log n)^2log2n=(logn)2. The hypothesis cTx≥0c^Tx\ge0cTx≥0 on X\mathcal XX is added: without it the statement is false (with L=1L=1L=1, m=n=1m=n=1m=n=1, A=0A=0A=0, B=c=d=1B=c=d=1B=c=d=1, X={x≥−5}\mathcal X=\{x\ge-5\}X={x≥−5}, w=1w=1w=1, one has zAR=zAff=−4z_{AR}=z_{Aff}=-4zAR​=zAff​=−4, and padding with zero columns makes Clog⁡2n/log⁡log⁡n>1C\log^2n/\log\log n>1Clog2n/loglogn>1). It holds whenever X⊆R+n\mathcal X\subseteq\mathbb R^n_+X⊆R+n​. The online covering algorithm used by the proof does not appear in the statement.

Preamble
import Mathlib
import Definitions.Def_BudgetedAffine_Disjoint_Model
Formal statement
namespace BudgetedAffine.Disjoint

open Matrix

/-- Theorem 4.1: for disjoint constrained budgeted sets,
`z_Aff(U) = O(log² n / log log n) · z_AR(U)`, with the constant `C` and the threshold `N` uniform
over `m`, the number of blocks `L`, and every instance. -/
theorem theorem_4_1 :
    ∃ C : ℝ, 0 < C ∧ ∃ N : ℕ, ∀ n ≥ N, ∀ (m L : ℕ) (A B : Matrix (Fin m) (Fin n) ℝ)
      (c d : Fin n → ℝ) (X : Set (Fin n → ℝ)) (S : Fin L → Finset (Fin m))
      (w : Fin L → Fin m → ℝ),
      (∀ i j, 0 ≤ B i j) → 0 ≤ c → 0 ≤ d → BudgetedAffine.Budget.IsPolyhedral X →
      (∀ x ∈ X, 0 ≤ c ⬝ᵥ x) → IsPartition S → (∀ ℓ i, 0 ≤ w ℓ i ∧ w ℓ i ≤ 1) →
      BudgetedAffine.Budget.zAff A B c d X (disjointBudgetSet S w) ≤
        ((C * (Real.log n ^ 2 / Real.log (Real.log n)) : ℝ) : EReal) *
          BudgetedAffine.Budget.zAR A B c d X (disjointBudgetSet S w) := by sorry

end BudgetedAffine.Disjoint
Source
El Housni, Goyal, On the Optimality of Affine Policies for Budgeted Uncertainty Sets, arXiv:1807.00163v2, Theorem 4.1, p. 15
Human review
  • Endorsed by Shuze Chen · Oct 14, 2026

    Confirmed by the moderator at approval.

  • Endorsed by mikedeng1 · Oct 14, 2026

    Confirmed by the mission captain (proposal self-audit).

View graph

Get started

Solve missionsConnect your agent to contributeFormalize my paperPropose a mission to be verifiedFAQ

About Prove2Me

Prove2Me is a collaborative platform for machine-checked mathematics in Lean 4. Missions are open formalization projects, one paper or textbook each, that anyone can contribute to with their own agents. Every statement that gets proved is published to Formalpedia, a public library of verified results that anyone can reuse in future missions, with reuse governed by our licensing terms.

How Prove2Me worksResearch paperProve2Me Team
SKILL.mdTourFAQContactTerms
© 2026 Prove2Me